Abstract

   graphic file with name ao3c00627_0014.jpg

   Aim: Growing period is an important period for fat remodeling. High-fat
   diet and exercise are reasons for adipose tissue (AT) remodeling, but
   existing evidence is not enough. Therefore, the effects of
   moderate-intensity continuous training (MICT) and high-intensity
   interval training (HIIT) on the proteomic characteristics of the
   subcutaneous AT of growing rats on normal diet or high-fat diet (HFD)
   were determined. Methods: Four-week-old male Sprague–Dawley rats (n =
   48) were subdivided into six groups: normal diet control group, normal
   diet–MICT group, normal diet–HIIT group, HFD control group, HFD–MICT
   group, and HFD–HIIT group. Rats in the training group ran on a
   treadmill 5 days a week for 8 weeks (MICT: 50 min at 60–70% VO[2max]
   intensity; HIIT: 7 min of warm-up and recovery at 70% VO[2max]
   intensity, 6 sets of 3 min of 30% VO[2max] followed by 3 min 90%
   VO[2max]). Following physical assessment, inguinal subcutaneous adipose
   tissue (sWAT) was collected for proteome analysis using tandem mass tag
   labeling. Results: MICT and HIIT attenuated body fat mass and lean body
   mass but did not affect weight gain. Proteomics revealed the impact of
   exercise on ribosome, spliceosome, and the pentose phosphate pathway.
   However, the effect was reversed on HFD and normal diet. The
   differentially expressed proteins (DEPs) affected by MICT were related
   to oxygen transport, ribosome, and spliceosome. In comparison, the DEPs
   affected by HIIT were related to oxygen transport, mitochondrial
   electron transport, and mitochondrion protein. In HFD, HIIT was more
   likely to cause changes in immune proteins than MICT. However, exercise
   did not seem to reverse the protein effects of HFD. Conclusion: The
   exercise stress response in the growing period was stronger but
   increased the energy metabolism and metabolism. MICT and HIIT can
   reduce fat, increase muscle percentage, and improve maximum oxygen
   uptake in rats fed with HFD. However, in rats with normal diet, MICT
   and HIIT triggered more immune responses of sWAT, especially HIIT. In
   addition, spliceosomes may be the key factors in AT remodeling
   triggered by exercise and diet.

1. Introduction

   The global prevalence of overweight and obesity is increasing
   remarkably and is also evident in children and adolescents.^[32]1
   High-fat diet (HFD) is an important risk factor for the disease and has
   greater temptation for children.^[33]2,[34]3 Thus, overweight and obese
   youths have much more risk of becoming overweight adults.^[35]4

   Increasing physical activity benefits the health of children and young
   people,^[36]5 and exercise plays an active role in the prevention and
   treatment of overweight and obesity.^[37]6 Moderate-intensity
   continuous training (MICT) and high-intensity interval training (HIIT)
   are the most common forms of exercise intervention to improve the
   various biological processes of multiple organs^[38]7−[39]9 and prevent
   obesity,^[40]10 but their effects and mechanisms are different.

   The body’s excess energy is stored in adipose tissue (AT) in the form
   of triglycerides. AT is considered one of the largest endocrine organs
   regulating obesity and other metabolic pathways.^[41]11 Therefore, AT
   is an important target organ for studying the occurrence of overweight
   and obesity. However, AT distribution and function in different body
   compartments are different. Subcutaneous white adipose tissue (sWAT)
   plays a fundamental role in systemic glucose homeostasis and muscle
   metabolism.^[42]12 Subcutaneous AT removal decreases glucose tolerance
   and insulin sensitivity in mice fed with HFD.^[43]13 Moreover,
   subcutaneous AT relocation to the visceral cavity improves glucose
   tolerance in obese mice induced by HFD.^[44]14 However, the angiogenic
   capacity of subcutaneous AT would be reduced by morbid obesity,
   resulting in insulin resistance and metabolic diseases.^[45]15 In a
   word, subcutaneous AT plays a key role in obesity regulation.

   HFD and exercise (including HIIT and MICT) are reasons for AT
   remodeling, and they have opposite effects.^[46]16−[47]18 Although
   exercise is recognized for its health benefits in overweight or obese
   children, whether exercise can reverse the health damage of HFD, what
   type of exercise can correct nutritional remodeling, and the molecular
   mechanism is still unclear, especially for growing children.

   Several studies have described the proteomic characteristics of white
   adipose tissue in mice or rats, but the high lipid content of adipose
   tissue, the difficulty of protein quantification, and the high number
   of biological repetitions of some experimental methods have limited the
   analysis of the overall protein changes in adipose tissue. We chose the
   method of TMT labeling to detect our samples. The 10-plex TMT method
   using synchronous parent ion scanning in Orbitrap Fusion is very
   accurate and able to measure small expression differences in biological
   samples. And this is a method that has been used to great effect by
   others.^[48]19−[49]21 Exercise habits developed early in life can carry
   over into adulthood, providing lifelong health benefits.^[50]22 Thus,
   the effects of 8-week MICT and HIIT on the proteomics of subcutaneous
   AT of growing rats in normal diet or HFD were observed to: (a) clarify
   the characteristics of MICT and HIIT on the proteomics of subcutaneous
   AT in growing rats in normal diet or HFD; (b) compare the effects of
   MICT or HIIT under different nutritional statuses on the proteomics of
   subcutaneous AT in growing rats; and (c) explore whether exercise could
   improve the metabolism of the subcutaneous AT of rats in HFD and which
   exercise model was much effective.

2. Materials and Methods

2.1. Animals

   The Animal Research Ethics Committee of Beijing Sport University
   (Beijing, China) approved the experiment protocols under license number
   201937A. Forty-eight 3-week-old specific pathogen-free (SPF) weaning
   Sprague–Dawley rats were purchased from Beijing Weitonglihua Laboratory
   Animal Co., Ltd., Beijing (License No. SCXK Beijing 2014-008). The
   animals were housed under standard SPF laboratory conditions (20–24 °C,
   40–60% relative humidity, light: dark 12:12 h) and provided with
   pelleted food and water ad libitum.

   After 1 week of adaptive feeding, the animals were randomly divided
   into normal diet control group (C, n = 8), normal diet–MICT group (CE,
   n = 8), normal diet–HIIT group (CH, n = 8), HFD control group (HC, n =
   8), HFD–MICT group (HE, n = 8), and HFD–HIIT group (HH, n = 8). Groups
   C, CE, and CH were fed with control feed (3.82 kJ/g; 10% energy from
   fat, 20% from protein, and 70% from carbohydrate; Beijing Huafukang
   Biotechnology Co. Ltd., LOT No: D12451B); groups HC, HE, and HH were
   fed with high-fat feed with 45 kcal% fat (4.7 kJ/g; 45% energy from
   fat, 20% from protein, and 35% from carbohydrate; Beijing Huafukang
   Biotechnology Co., Ltd., LOT No: D12451) using the same formulation
   with D12451 from Research Diets.

2.2. Exercise Intervention

   The training group received the exercise intervention for 8 weeks, 5
   days a week. Rats of the CE and HE groups underwent 60–70% VO[2max]
   intensity continuous treadmill training for 50 min. Before and after
   each training, the CH and HH groups were given 7 min each of warm-up
   and recovery treadmill training at 70%VO[2max] intensity, followed by
   six sets of HIIT (3 min at 90%VO[2max] + 3 min at 50%VO[2max]) for a
   total of 50 min. VO[2max] tests were performed every 2 weeks to
   calibrate speed ([51]Figure [52]1
   ).

Figure 1.

   [53]Figure 1
   [54]Open in a new tab

   Experiment process. (A) Exercise intervention, (B) experiment process.

2.3. Sample Selection

   Sample collection was performed 48 h after the last exercise to avoid
   the effects of acute exercise. Dexa testing (XR-46, Norland) of body
   weight was performed prior to lethality. Rats were anesthetized with
   10% chloral hydrate and then sacrificed by blood sampling from the
   abdominal aorta. Bilateral inguinal subcutaneous AT was collected and
   stored at −80 °C for subsequent testing. Finally, three rats from each
   group who completed the intervention process were randomly selected for
   tandem mass tag (TMT) proteomics testing.

2.4. TMT Proteomics

2.4.1. Protein Extraction and Digestion

   The sample was dissolved in lysis buffer (7 M urea [Bio-Rad], 2 M
   thiourea [Sigma-Aldrich], and 0.1% CHAPS [Bio-Rad]). Then, the tissue
   was ground with three titanium dioxide abrasive beads (70 Hz, 120 s)
   followed by centrifugation at 5000g for 5 min at 4 °C. The supernatant
   was collected and centrifuged at 15,000g for 30 min at 4 °C. The final
   supernatant was collected and stored at −80 °C until used. Total
   protein concentration was measured by the Bradford protein assay. Total
   protein (200 μg) from each sample was incubated at 55 °C for 1 h with 5
   μL of 200 mM reducing reagent. Then, 5 μL of 375 mM iodoacetamide was
   added and incubated for 10 min at room temperature (RT) in the dark.
   Next, 200 μL of 100 mM dissolution buffer (AB Sciex) was added and
   centrifuged at 12,000g for 20 min. The sample was digested in trypsin
   for 14 h at RT, lyophilized, and redissolved with 100 mM dissolution
   buffer for labeling.

2.4.2. TMT Labeling

   The TMT10plex reagent (Thermo Scientific, cat.#90111) was incubated at
   RT. Absolute ethyl alcohol (41 μL) was added to the TMT reagent (0.8
   mg/tube) and mixed well. Because there were more samples than could be
   multiplexed in one TMT set, balanced batches were produced, and a
   common pool of all samples was also labeled and added to each batch.
   The TMT reagent (41 μL) was added to 100 μg of the digested sample.
   Then, the mixture was oscillated, centrifuged, incubated for 1 h at RT,
   and incubated in 5% quenching reagent (8 μL) for 15 min to terminate
   the reaction. The samples were stored after lyophilization. The
   efficiency of chemical tagging was 95%. TMT-labeled peptides were
   fractionated via basic reversed-phase chromatography on a RIGOL L-3000
   dual-gradient high-performance liquid chromatograph (HPLC) using an
   Agela Durashell-C18 column (4.6 mm × 250 mm i.d., 5 μm, 100 Å). The
   column oven was set at 45 °C. Prior to loading peptides, the C18 column
   was washed with 100% methanol and equilibrated with Buffer A (0.1%
   ammonium hydroxide and 2% acetonitrile). Peptides were injected and
   eluted from the column using a gradient of mobile phase A (2% ACN, 0.1%
   NH[4]OH) to mobile phase B (98% ACN, 0.1% NH[4]OH) over 72 min at a
   flow rate of 0.7 mL/minute. The solvent gradient was set as follows:
   5%B, 0 min; 5–8% B, 5 min; 8–18% B, 30 min; 18–32% B, 27 min; 32–95% B,
   6 min; 95–5%, 4 min. The fractions collected were orthogonally
   concatenated into 10 pooled fractions.

2.4.3. Peptide Identification by Nano–Ultrahigh-Performance Liquid
Chromatography (UPLC)–Tandem Mass Spectrometry (MS/MS)

   The acquired peptide fractions were suspended with 20 μL of buffer A
   (0.1% formic acid, 2% acetonitrile) and centrifuged at 12,000 rpm for
   10 min. The supernatants (10 μL) were injected into the nano-UPLC–MS/MS
   system consisting of a Nanoflow high-performance liquid chromatography
   system (EASY-nLC 1000, Thermo Scientific) and an Orbitrap Fusion Lumos
   mass spectrometer (Thermo Scientific). The sample was loaded onto an
   Acclaim PepMap100 C18 column and then separated by an EASY-Spray C18
   column (Beijing Qinglian Biotech Co., Ltd, China). Peptides were
   separated using a gradient from 6 to 12% B in 11 min, then 12–20% B in
   37 min, and stepped up to 32% in 20 min followed by a 7 min wash at 95%
   B at 600 nL/min, where solvent A was 0.1% formic acid in water and
   solvent B was 80% ACN and 0.1% formic acid in water. The total duration
   of the run was 75 min. The mass spectrometer was operated in
   positive-ion mode (2.0 kV source voltage), and full MS scans were
   performed in the Orbitrap over the range of 300–1500 m/z at a
   resolution of 120,000 with an automatic gain control (AGC) target of
   4e^5 and a maximum ion injection time of 50 ms. The scan range was
   350–1550 m/z. Following higher-energy collisional dissociation (HCD)
   with a normalized collision energy (NCE) of 32%, MS/MS spectra were
   collected in the Orbitrap (60,000 resolution) with an AGC target of
   5e^4 and a maximum ion injection time of 118 ms. The database used in
   this experiment is the Uniprot_RAT database. The resulting MS/MS data
   were processed using Proteome Discoverer 2.1.

2.4.4. Protein Identification

   Identification was set as follows: fully tryptic specificity, maximum
   of two missed cleavages, minimum peptide length of 6, fixed
   modifications for TMT tags on lysine residues and peptide N-termini
   (+229.162932Da), and carbamidomethylation of cysteine residues
   (+57.02146Da), variable modifications for oxidation of methionine
   residues (+15.99492Da), precursor mass tolerance of 15 ppm and a
   fragment mass tolerance of 0.02 Da for MS2 spectra collected in the
   Orbitrap. A percolator was used to filter peptide spectral matches and
   peptides to a false discovery rate (FDR) of less than 1%. After
   spectral assignment, peptides were assembled into proteins and were
   further filtered based on the combined probabilities of their
   constituent peptides to a final FDR of 1%. Only unique and razor
   peptides were considered for quantification.

2.4.5. Bioinformatics Analysis

   C2 gene sets from Molecular Signatures Database Collections
   (c2.cp.kegg.v7.5.symbols.gmt) identified the proteome signatures
   enriched in CE versus C and CH versus C using the Gene Set Enrichment
   Analysis (GSEA) software (v4.2.1). The criteria of differentially
   expressed proteins (DEPs) were fold change >1.5 or <0.67, and P < 0.05
   was considered statistically significant. In this study, the Gene
   Ontology (GO) terms and Kyoto Encyclopedia of Genes and Genomes (KEGG)
   pathway enrichment analysis of DEPs were automatically completed and
   visualized by clusterProfiler version 4.3.0 in R software.

2.5. Western Blot Validation

   Total proteins were extracted with RIPA buffer, and the protein
   concentration was determined using the BCA protein assay kit. Equal
   amounts of protein (20–40 μg) were resolved on SDS-PAGE gels and then
   transferred to PVDF membranes. The following antibodies were used:
   β-actin (mouse antibody; 1:2000 dilution; CWbio, Taizhou, China); Pfkp
   (rabbit antibody; 1:500 dilution; Proteintech, Wuhan, China); Crp
   (rabbit antibody; 1:2000 dilution; Proteintech, Wuhan, China); Fasn
   (rabbit antibody; 1:1000 dilution; Cell Signaling Technology, Boston).
   The membranes were then incubated with anti-rabbit or anti-mouse IgG
   HRP-conjugated secondary antibodies. The bands were visualized with
   enhanced chemiluminescence (ECL) reagents and analyzed using the
   Chemi-Doc System (Bio-Rad). The levels of target proteins were
   normalized against the β-actin protein levels. Band intensities were
   measured using ImageJ software (NIH).

2.6. Statistical Analyses

   Statistical analyses were performed in SPSS 22.0 and visualized in
   GraphPad Prism version 8.0 (GraphPad Software, San Diego, CA).
   Two-factor ANOVA was used for body weight, body fat percentage, lean
   body mass, skeletal muscle percentage, and maximum oxygen uptake. The
   material in [55]Figure [56]1
   is from [57]https://smart.servier.com/.

3. Results

3.1. Physical Characteristics of Animals

   Before the start of the experiment, we ensured that the body weights of
   rats in each group had no remarkable differences ([58]Figure [59]2
   A). After 8 weeks of intervention, the HFD-fed rats showed a much
   poorer haircoat quality. MICT and HIIT reduced the body weight of
   normal-diet-fed rats (P < 0.01), but no significant differences were
   found in HFD-fed rats (P > 0.05). However, as expected, the body fat
   percentages of rats in the HE and HH groups were significantly lower
   than that in the HC group (P < 0.01), and the lean body weight (P <
   0.05) and the body muscle percentage of (P < 0.01) in the HE and HH
   groups were significantly higher than those in the HC group ([60]Figure
   [61]2B–D). In addition, the wet weight of subcutaneous AT was also
   measured. The tissue weight in the HE group was remarkably higher than
   that in the CE group, and surprisingly, the subcutaneous AT of rats in
   the HE group was heavier than that in the HC group ([62]Figure [63]2E).
   Before the intervention, the VO[2max] values of rats in different
   groups had no remarkable differences. After the intervention, the
   VO[2max] in the exercise groups was higher than that in the control
   groups under the same diet intervention. Only the HFD group had a
   statistically different VO[2max]. The HFD group had lower VO[2max] than
   the normal diet group even with MICT.

Figure 2.

   [64]Figure 2
   [65]Open in a new tab

   Physical characteristics of animals. (A) Body weight, (B) body fat
   percentage, (C) body muscle percentage, (D) lean body mass, (E)
   bilateral inguinal subcutaneous AT wet weight, (F) maximum oxygen
   uptake. ^aP < 0.05, compared with C group; ^bP < 0.05, ^bbP < 0.01,
   compared with CE group; ^dP < 0.05, ^ddP < 0.01, compared with HC
   group; ^eP < 0.05, ^eeP < 0.01 compared with HE group.

3.2. GSEA Reveals the Effects of MICT and HIIT on AT

   A total of 4265 individual proteins were identified and quantified in
   total in this study. GSEA was carried out to identify the effects and
   mechanisms of different exercise modes on AT. The analysis indicated
   that the dataset has 3645 genes for CE versus C comparison. A total of
   1953 highly expressed genes were found in the CE group with a
   correlation area of 58.0%, and 1692 genes were highly expressed in the
   C group with a correlation area of 42.0%. The most significantly
   enriched gene sets positively correlated with the CE group included
   ribosome, spliceosome, pentose phosphate pathway (PPP), Fc γ R-mediated
   phagocytosis, and ubiquitin-mediated proteolysis (P < 0.05). The most
   significantly enriched gene sets negatively correlated with the CE
   group included PPAR signaling pathway, N-glycan biosynthesis, and
   lysosome (P < 0.05, [66]Figure [67]3
   A,B).

Figure 3.

   [68]Figure 3
   [69]Open in a new tab

   GSEA enrichment in the normal diet group. (A) Left is a heatmap of the
   top 50 proteins in CE versus C. Red means upregulation, and blue means
   downregulation. (B) Pathways with significant differences in CE versus
   C. (C) Left is a heatmap of the top 50 proteins in CH versus C. (D)
   Pathways with significant differences in CH versus C.

   A total of 3645 genes were identified in the CH versus C comparison,
   1916 genes were highly expressed in the CH group with a correlation
   area of 52.7%, and 1729 genes were highly expressed in the C group with
   a correlation area of 47.3%. The most significantly enriched gene sets
   positively correlated with the CH group included ribosome, proteasome,
   spliceosome, aminoacyl tRNA biosynthesis, pentose phosphate pathway,
   citrate cycle, and tricarboxylic acid cycle (P < 0.05). The most
   significantly enriched gene sets negatively correlated with the CH
   group included complement and coagulation cascades, extracellular
   matrix (ECM)–receptor interaction, and leishmania infection (P < 0.05,
   [70]Figure [71]3
   C,D).

   The analysis indicated that the dataset has 3645 genes for the HE
   versus HC comparison. A total of 1760 highly expressed genes were found
   in the HE group with a correlation area of 44.9%, and 1885 genes were
   highly expressed in the HC group with a correlation area of 55.1%. The
   most significantly enriched gene sets positively correlated with the HE
   group included complement and coagulation cascades, gap junction, snare
   interactions in vesicular transport, N-glycan biosynthesis, and
   endocytosis (P < 0.05). The most significantly enriched gene sets
   negatively correlated with the HE group included spliceosome,
   peroxisome, proteasome, ribosome, and cysteine and methionine
   metabolism (P < 0.05, [72]Figure [73]4
   A,B).

Figure 4.

   [74]Figure 4
   [75]Open in a new tab

   GSEA enrichment in the HFD group. (A) Heatmap of the top 50 proteins in
   HE versus HC. (B) Pathways with significant differences in HE versus
   HC. (C) Heatmap of the top 50 proteins in HH vs HC. (D) Pathways with
   significant differences in HH versus HC.

   A total of 3645 genes were also obtained in the HH versus HC
   comparison. A total of 1509 genes were highly expressed in the HH group
   with a correlation area of 42.7%, and 2136 genes were highly expressed
   in the HC group with a correlation area of 57.3%. The most
   significantly enriched gene sets positively correlated with the HH
   group included complement and coagulation cascades, drug metabolism
   cytochrome p450, N-glycan biosynthesis, ECM–receptor interaction,
   glycerophospholipid metabolism, and chemokine signaling pathway (P <
   0.05). The most significantly enriched gene sets negatively correlated
   with the HH group included proteasome, spliceosome, lysosome, aminoacyl
   tRNA biosynthesis, pyruvate metabolism, valine, leucine and isoleucine
   degradation, pyrimidine metabolism, and cysteine and methionine
   metabolism (P < 0.05, [76]Figure [77]4
   C,D).

3.3. Functional Correlation Analysis of DEPs in a Normal Diet

   A total of 139 proteins were differentially abundant between groups,
   among which 88 were upregulated and 51 were downregulated in the CE
   group compared with C group sWAT ([78]Figure [79]5
   A). GO function annotation analysis showed that the DEPs affected by
   MICT were related to hemoglobin complex, which influenced oxygen
   transport and oxygen carrier activity ([80]Figure [81]5C). KEGG also
   showed that spliceosome and lysosome were enriched remarkably
   ([82]Figure [83]5B). The DEPs were sorted according to P value. The
   most significantly altered proteins in CE versus C were involved in the
   neuromuscular process (Mycbp2, Gba) and endoplasmic reticulum (Gba,
   Lnpk, S100a10, Syvn1; [84]Table S1). The maximal clique centrality
   (MCC) method of the CytoHubba plug-in in Cytoscape was used to screen
   the top 10 hub genes. The hub genes were associated with spliceosome
   (Srsf 2, Srsf 5, Srsf 6, and Srsf 11; [85]Figure [86]5D).

Figure 5.

   [87]Figure 5
   [88]Open in a new tab

   Functional correlation analysis of DEPs in CE versus C. (A) Volcano map
   of DEPs in CE versus C. (B) KEGG analysis in CE versus C. The x-axis
   label represents gene ratio, and the y-axis label represents the KEGG
   pathway. The size of the dot indicates the amount of genes under a
   specific term, and the color of the dot indicates the P-value. (C) GO
   function annotation analysis in CE versus C. (D) Top 10 hub genes of
   PPI in CE versus C. The redder the color, the higher the MCC score.

   A total of 204 DEPs (99 upregulated and 105 downregulated) were
   identified in the CH group compared with the C group ([89]Figure [90]6
   A). GO functional annotation analysis showed that the DEPs affected by
   HIIT were related to lipoprotein particle, as well as biological
   processes and molecular functions involved in oxygen transport and
   oxygen carrier activity ([91]Figure [92]6C). KEGG also showed that
   lysosomes were enriched substantially ([93]Figure [94]6B). The most
   remarkably altered proteins in CH versus C were involved in
   mitochondrial electron transport (Ndufs8, Coq9), lysosome organization
   (Tmem106b, Gba), and mitochondrion protein (Cmc2, [95]Table S2). The
   hub genes were associated with apolipoprotein ([96]Figure [97]6D).

Figure 6.

   [98]Figure 6
   [99]Open in a new tab

   Functional correlation analysis of DEPs in CH versus C. (A) Volcano map
   of DEPs in CH versus C. (B) KEGG analysis in CH versus C. (C) GO
   function annotation analysis in CH versus C. (D) Top 10 hub genes of
   PPI in CH versus C. The redder the color, the higher the MCC score.

3.4. Functional Correlation Analysis of DEPs in an HFD

   Five DEPs were downregulated in the HE group compared with the HC group
   ([100]Figure [101]7
   A). The enriched GO terms included mitochondrial electron transport,
   aerobic electron transport chain, and aerobic respiration ([102]Figure
   [103]7B). The most remarkably altered proteins in HE versus HC were
   involved in ATPase activity (Hspe1, Tor3a), mitochondrial electron
   transport (Cox6c2), and cell adhesion (Cd44, [104]Table S3).

Figure 7.

   [105]Figure 7
   [106]Open in a new tab

   Functional correlation analysis of DEPs in HE versus HC. (A) Volcano
   map of DEPs in HE versus HC. (B) GO function annotation analysis in HE
   versus HC.

   A total of 71 upregulated and 17 downregulated DEPs were identified in
   the HH group compared with the HC group ([107]Figure [108]8
   A). GO function annotation analysis showed that the DEPs affected by
   HIIT were related to Golgi ([109]Figure [110]8C). KEGG also showed that
   lysosomes were enriched remarkably ([111]Figure [112]8B). The most
   substantially altered proteins in HH versus HC were involved in immune
   response (Ig-like domain-containing protein), regulation of calcium ion
   transport (Cacna2d1, Ank2), and autophagosome (Tbc1d12, Atg12;
   [113]Table S4). Unfortunately, many hub gene functions still need to be
   explored.

Figure 8.

   [114]Figure 8
   [115]Open in a new tab

   Functional correlation analysis of DEPs in HH versus HC. (A) Volcano
   map of DEPs in HH versus HC. (B) KEGG analysis in HH versus HC. (C) GO
   function annotation analysis in HH versus HC. (D) Top 10 hub genes of
   PPI in HH versus HC. The redder the color, the higher the MCC score.

3.5. Response of Different Nutritional States to Exercise

   The nutrition-induced effects of exercise on the proteome of AT were
   determined. A total of 139 and 5 DEPs were identified in CE versus C
   and HE versus HC, respectively. However, no common DEP was found
   between the two kinds of states ([116]Figure [117]9
   A).

Figure 9.

   [118]Figure 9
   [119]Open in a new tab

   Differential effects of exercise in two nutritional states. (A) Venn
   diagram of CE versus C and HE versus HC. (B) Venn diagram of CH versus
   C and HH versus HC. (C) Heatmap of DEPs in CH versus C and HH vs HC.
   (D) Heatmap of common DEPs in CH versus C and HH versus HC, including
   22 proteins. (E) GO terms enriched in CH versus C except 22 common
   DEPs. (F) GO terms enriched in HH versus HC except 22 common DEPs. (G)
   GO terms enriched in 22 common DEPs.

   A total of 204 and 88 DEPs were identified in CH versus C and HH versus
   HC, respectively. The two nutritional states had 22 common DEPs
   ([120]Figure [121]9
   B). Functional analysis showed that the 22 common DEPs were enriched in
   the GO terms, pyridoxal phosphate binding, maintenance of location,
   maintenance of mitochondrion location, and respiratory electron
   transport chain, and 182 (except 22 common proteins) DEPs in CH versus
   C were enriched in the GO terms, phospholipid efflux, oxygen transport,
   and oxygen carrier activity. Sixty-six DEPs (except 22 common proteins)
   in HH versus HC were enriched in the GO terms, positive regulation of
   autophagy and RNA polymerase II.

3.6. Exercise Training Did Not Reverse the Effects of HFD on AT

   A total of 27 DEPs were found in HC versus C, including seven
   upregulated and seven downregulated proteins. The enrichment analyses
   of GO and KEGG showed that HFD caused the abnormal expression of lipid
   metabolism proteins, such as those related to cholesterol metabolism,
   fat digestion, and absorption ([122]Figure [123]10
   ). The related downregulated proteins included Apoa1, Apoa2, and Apoa4.
   MICT did not reverse the trend of protein expression induced by HFD.
   HIIT reversed the trend of nine protein expression induced by HFD, and
   six Ig-like domain-containing proteins were upregulated.

Figure 10.

   [124]Figure 10
   [125]Open in a new tab

   Functional correlation analysis of DEPs in HC versus C. (A) Volcano map
   of DEPs in HC versus C. (B) KEGG analysis in HC versus C. (C) GO
   function annotation analysis in HC versus C.

3.7. Protein Expression Levels of Enrichment Pathways in Different Groups

   Based on the enrichment analysis results of GSEA and DEPs, we have
   listed the original data of some proteins. The following figure shows
   some key results ([126]Figure [127]11
   ).

Figure 11.

   [128]Figure 11
   [129]Open in a new tab

   Protein expression levels of enrichment pathways in different groups.
   (A) Hemoglobin proteins, (B) inflammation, (C) spliceosome, (D)
   oxidoreductase, (E) pentose phosphate pathway, (F) mitochondrion
   proteins.

3.8. Proteomics Validation

   To confirm whether differently expressed proteins were consistent with
   that determined by TMT-LC-MS/MS, we identified 3 of them ([130]Figure
   [131]12
   ). The relative expression levels of all of the proteins quantified by
   Western blot followed the trend of expression levels and significance
   determined by TMT quantification.

Figure 12.

   [132]Figure 12
   [133]Open in a new tab

   Results of Western blot. (A) Pfkp, (B) Crp, (C) Fasn.

4. Discussion

   Although many proteomic studies on the effects of exercise on the body
   have been conducted, AT research is still scarce. A complete and
   comprehensive overview of the field is lacking because of the great
   variability of animal models, training programs, and other factors. AT
   is an important organ that reflects energy balance, and diet and
   exercise are important factors affecting energy balance.^[134]23
   Therefore, we carried out a study of the effects of diet and exercise
   on the proteomic characteristics of AT.

   We observed that exercise from childhood promoted body skeletal muscle
   percentage and reduced body weight. HIIT and MICT activated ribosome,
   spliceosome, and the pentose phosphate pathway. Spliceosome, which
   affects RNA processing, plays an important role in protein diversity
   during development and adaptation.^[135]24,[136]25 It also affects AT
   production and thus has an effect on obesity; for example, knockdown of
   SRSF10 and SRSF2 adversely affects adipogenesis.^[137]26−[138]28 Our
   results showed that most splicing factors were increased in the
   exercise group, including SRSF10 and SRSF2, which might mean that the
   reduction in fat percentage could be detrimental. It may also be the
   body’s resistance to stress response.^[139]29 Physical activity
   affecting the spliceosome is recognized, but most studies are related
   to skeletal muscle.^[140]30 Aldiss et al.^[141]31 showed that rats
   swimming in thermoneutral water have a decreased expression of
   spliceosomes. Inconsistencies among training temperatures, training
   objects, and training programs may lead to inconsistent results. In any
   case, the role of the spliceosome in the influence of exercise on the
   body should be the direction of future research. A study of
   transcriptional profile in AT from healthy men with 6-month supervised
   exercise intervention identified pathways enriched in response to
   exercise, including the ribosome and proteasome.^[142]32 Aldiss et al.
   also found that ribosomes play an important role in white AT through
   exercise training, but studies on related functions are still few. In
   addition, we need to acknowledge the importance of ribosomes in the
   field of sports, as they may compete with mitochondria in skeletal
   muscles.^[143]33 The pentose phosphate pathway is a fundamental
   component of cellular metabolism. It is important for maintaining
   carbon homeostasis, providing precursors for nucleotide and amino acid
   biosynthesis, providing reducing molecules for anabolism, and combating
   oxidative stress. DEPs showed that both modes of exercise were related
   to oxygen transport, and the expression of glycolytic protein Pfkp was
   significantly increased. HIIT involved the expression of HIF-1
   pathway-related proteins and hypoxia-response-related pathways, which
   also suggested that HIIT is involved in the hypoxia-inducible
   environment. Some studies have shown that HIIT triggers the HIF-1
   pathway in skeletal muscle,^[144]34,[145]35 but the effect on AT is
   still less. Our results showed that exercise boosts the body’s
   metabolism.^[146]36 Although the AT proteome and the ability to lose
   fat were somewhat similar between the two exercise modes, the proteomes
   affected by HIIT appeared to be concentrated in energy metabolism and
   biosynthesis, whereas the proteome affected by MICT was involved in the
   physiological function of protein and redox. Overall, exercise is
   beneficial because C-reactive protein (CRP) all declined after
   exercise, and its high expression is highly correlated with chronic
   metabolic diseases such as cardiovascular disease and diabetes.^[147]37

   However, we found interesting results in HFD-fed rats that underwent
   exercise, which were reversed from the normal diet group. For example,
   opposite results existed for the expression of spliceosomes. In the
   case of HFD, a negative correlation exists between spliceosome
   expression and exercise. Therefore, attention should be given to the
   effect of exercise on spliceosome. In HFD, the molecular response to
   HIIT seems to be stronger than that to MICT and affected more metabolic
   pathways and immune responses, although both modes of exercise can play
   a role in reducing fat and building muscle. Surprisingly, sWAT wet
   weight increased considerably when MICT was performed. Body fat
   percentage decreased; therefore, we reexamined the visceral AT data and
   found a remarkable decrease in visceral AT. Therefore, whether the
   increased sWAT content caused by MICT is an increase of “good fat”
   needs to be further explored.

   In addition, the focus of our study was to understand the effects of
   exercise at different nutritional levels. Compared with normal diet,
   MICT with HFD affected the expression of a variety of mitochondrial
   proteins, whereas HIIT produced more immune stress proteins (about 35
   kinds), such as immunoglobulin complex. During MICT, we did not find
   the same DEPs between the two nutritional states, suggesting that MICT
   responds differently to nutrition. Under normal diet, MICT affected
   various physiological processes and downregulated various hemoglobin
   genes (Hba1, Hba-a3, and Hbb) related to oxygen utilization. This
   result seems to be surprising. Whether the result means an increase in
   anaerobic metabolism still needs to be confirmed and explored in other
   tissues and organs. Some metabolism-related proteins that were
   downregulated by MICT in a high-fat state aroused our interest. For
   example, CD44 expression is elevated in obesity-induced white AT
   inflammation, and CD44 k/o mice are phenotypically healthier.^[148]38
   Our experiments observed that MICT with HFD led to a decrease in CD44,
   which seems to be a good performance. The cellular response to protein
   destruction is completed by the activation of heat shock
   proteins.^[149]39 The E1 heat shock protein family (Hspe1) encodes
   Hspup0 protein, which is an important part of the Hsp60/Hspup0 system
   that assists in the folding of proteins in the mitochondrial matrix
   space.^[150]40,[151]41 HSPE1 in AT can be modulated by the neutrophil
   extracellular traps in obesity to increase inflammation.^[152]42 Cox6c2
   is the component of cytochrome c oxidase, the last enzyme in the
   mitochondrial electron transport chain that drives oxidative
   phosphorylation. Liver Cox6c2 decreases under a high-fat and
   high-cholesterol diet.^[153]43 Although no AT-related data were
   obtained, the decrease in this protein caused by MICT with HFD needs to
   be considered.

   When HIIT was performed, 22 proteins existed in both nutritional
   states. Although one protein whose expression trend was opposite, we
   speculated that the 21 proteins might be the key proteins of HIIT
   affecting AT remodeling and the expression trend of these proteins was
   not affected by diet. Some of these proteins were involved in the
   respiratory electron transport chain, such as Coq9. Coenzyme Q, also
   known as ubiquinone, is a critical component of the mitochondrial
   electron transport chain. Coq9, which belongs to coenzyme Q, is a
   lipid-binding protein that can interact with Coq7 to enable coenzyme Q
   biosynthesis,^[154]44 which affects cellular respiration and
   contributes to cellular anti-oxidation.^[155]45 However, the role of
   this protein in AT still needs to be explored. The interacting proteins
   Cmc1 and Cmc2 act as a cytochrome c oxidase biogenesis factor to
   regulate mitochondrial superoxide dismutase activity, which plays an
   important role in cellular respiration and free radical
   scavenging.^[156]46,[157]47 A study^[158]46 showed that the absence of
   Cmc2 induces a 5-fold increase in Cmc1 accumulation in yeast cell
   mitochondrial membranes. However, whether exercise causes the opposite
   expression of these two proteins in our study remains to be explored.
   Glutathione transferase omega (Gsto1), which is expressed oppositely in
   the two nutritional states, plays a pro-inflammatory role. A study
   showed that Gsto1–/– mice are less susceptible to HFD-induced obesity
   and AT inflammation.^[159]48 However, in our study, Gsto1 was
   remarkably elevated in HIIT with HFD, suggesting a possible
   inflammatory response. Moreover, we analyzed the remaining proteins in
   each state except 22 proteins. The results showed that HIIT with normal
   diet also had the same decrease in hemoglobin genes (Hba-a1, Hba-a3,
   and Hbb) as MICT. We also noticed the downregulation of some
   lipoproteins, such as Apoa1, Apoa4, and Apom. This finding suggests
   that they are all beneficial circulating factors, which are also
   beneficial for obesity, insulin sensitivity, and
   atherosclerosis.^[160]49−[161]53 The result gives us a new hint that
   our HIIT regimen may have a certain negative impact on the molecular
   mechanisms of the organism, although we have obtained positive
   phenotypic effects. HIIT with HFD affected the positive regulation of
   autophagy, which upregulated Ras-related protein Rabup2 and two-pore
   calcium channel protein up (Tpc1). Rabup2 regulates
   mammalian/mechanistic target of rapamycin complex 1 activity and
   autophagy through the trafficking of PAT4, the initiation link of
   autophagy.^[162]54 TPC1, which belongs to two-pore channels,
   participates in the regulation of multiple endolysosomal trafficking
   pathways; its knockdown can cause lysosomal dysfunction and affect
   metabolism.^[163]55 Although HIIT resulted in positive phenotypic
   changes and some modest benefits at the molecular level in HFD, the
   resulting immune response was too strong. Whether this effect is
   beneficial still needs to be further explored in the future.

   We observed that an HFD for 8 weeks resulted in changes in lipid
   metabolism and a remarkable increase in body fat percentage; therefore,
   we investigated whether exercise could reverse the adverse effects of
   HFD. From the proteomic level, we did not observe that exercise
   reversed the changes in protein expression induced by HFD, and an
   increase in inflammation occurred during HIIT. However, in our study,
   we found individual differences in the effects of HFD on the body,
   which also led to some outliers in proteomic data. Such results are
   also related to the exercise intensity we chose. In spite of these
   factors, we still advocate a reasonable diet and exercise from the
   growth period to the maturity period according to the results.

   We acknowledge that this work has several limitations. First, our
   sample size needs to be expanded, and further verification has not been
   done. Second, our exercise program is still relatively single, and HIIT
   has a variety of combinations; therefore, future studies should also
   explore HIIT with different intensities to obtain more comprehensive
   and objective data.

5. Conclusions and Suggestions

   The following conclusions were drawn from the results. The exercise
   stress response in the growing period was stronger and increased energy
   metabolism and metabolism, suggesting that exercise during the growing
   period should be appropriate. The two modes of exercise can reduce fat,
   increase muscle percentage, and improve maximum oxygen uptake in rats
   fed with HFD and trigger more immune responses in sWAT compared with
   rats with normal diet. Therefore, a reasonable diet and proper exercise
   should still be implemented. MICT is less likely to elicit the immune
   response of sWAT than HIIT in high-fat conditions. Spliceosomes may be
   the key factors in AT remodeling triggered by exercise and diet.
   Exercise may not fully reverse HFD-induced changes in protein
   expression. Although exercise is phenotypically beneficial, we advocate
   proper diet and exercise in the growth period.

   Furthermore, our conclusions are not absolute. In the future, we still
   need to expand the sample size and change the exercise mode and
   intensity to further verify the results.

Acknowledgments